X

latest videos

What It's Like Using a Brain Implant With ChatGPT
 • Computing
connieguglielmoselfie-centered
meet connie

Connie Guglielmo is CNET's head of AI editorial strategy. She puts the spotlight on the latest and most important developments in artificial intelligence.

Full bio
glossary

artificial general intelligence, or AGI

A concept that suggests a more advanced version of AI than we know today, one that can perform tasks much better than humans while also teaching and advancing its own capabilities.

faqs

What is GPT-4o and how does it relate to ChatGPT?

GPT-4o (the o is for "omni") is the May 2024 update from OpenAI to the AI engine that powers ChatGPT, and it's two times faster than GPT-4 Turbo, which debuted in late 2023. Your prompts can include any combination of text, audio, images and video, and the responses generated will be able to include similar combinations. According to OpenAI, GPT-4o brings "GPT-4-level intelligence to everything" including the free version of ChatGPT, though paying customers will get five times the capacity. New real-time speech functionality means you can interact with ChatGPT in a more conversational way.

Learn more

What are AI tokens and how do they work?

Tokens are pieces of words that an LLM evaluates to understand the broader context of a query. Each token is made up of roughly four characters in English. They can be letters, numbers, spaces, special characters and more. For example, the sentence "You miss 100% of the shots you don't take" consists of 11 tokens. (If you swap out the percentage symbol for the word "percent," the count increases to 13 tokens.) Tokens are used both as inputs and outputs. When you input a query into an AI model, the model breaks down the words into tokens, analyzes them and delivers a response in tokens that are then converted into words you'll understand.

Learn more

What is an AI hallucination?

A generative AI model "hallucinates" when it delivers false or misleading information. This happens because of how the models are trained, on massive amounts of data, like articles, books, code and social media posts. In creating its response to a prompt, the AI makes inferences about what words should come next, but it lacks contextual information so it may generate wording that seems plausible but isn't true. Hallucinations can also result from improper training and/or biased or insufficient data, which leave the model unprepared to answer certain questions.

Learn more

What is Google's AI Overviews feature?

AI Overviews is a new spin on Google search. Instead of having to ask multiple questions and then getting multiple search results pages, you can ask something more complex up front. In response to your query, generative AI presents a tidy summation, researched and written in seconds. It also provides links to its sources for the answer, and Google's traditional links appear below that. Think of it as a sort of TL;DR for your search.

Learn more
Showing 1 of 4

our ai experts

Our writers and editors know the subject cold. Here are CNET's leading authorities on AI.

ll-headshot

Lisa Lacy

Lisa is CNET's lead AI writer, homing in on the latest developments and what you need to know now.

imad-khan

Imad Khan

Imad covers Google and all its AI products, as well as those from Microsoft, OpenAI and more.

lisa-eadicicco-profile-photo

Lisa Eadicicco

An expert in mobile tech, Lisa has the lowdown on how AI is showing up in the phones in your pocket.

katelyn-chedraoui-headshot

Katelyn Chedraoui

Katelyn writes about social media and online services, with an eye for how AI can work for you.

AI Atlas is 100% human-generated. For more, see CNET's AI policy.